SYSTEMATICS shalontaki, a New Cryptic Species of Ground (: : ) from the Southwestern United States

1 W. EVAN BRASWELL, LEANNA M. BIRGE, AND DANIEL J. HOWARD

Department of Biology, New Mexico State University, Las Cruces, NM 88003

Ann. Entomol. Soc. Am. 99(3): 449Ð456 (2006) ABSTRACT A new species of ground cricket in the Allonemobius (Hebard) is described from the southwestern United States. Like many singing Orthoptera, this species is morphologically cryptic and most easily distinguished from related species by using molecular characters and male calling song traits. We describe Allonemobius shalontaki as a new species and present data suggesting its repro- ductive isolation from congeneric species. Allonemobius shalontaki exhibits a unique male calling song and possesses a distinct allozyme genotype. Additionally, we present an electrophoretic key to the genus and propose phylogenetic relationships among species.

KEY WORDS Allonemobius, calling song, Nemobiinae, Orthoptera, phylogeny

Among the singing Orthoptera many closely related, ferentiation (Chu and Howard 1998, Mousseau and reproductively isolated species are divergent in eco- Howard 1998, Roff and Mousseau 1999, Roff et al. logical and sexual traits yet exhibit little or no mor- 1999), life history evolution (Tanaka and Brookes phological divergence. Species with these character- 1983, Bradford and Roff 1995, Carrie´re et al. 1997, Roff istics are known as cryptic species and pose particular and Bradford 2000, Olvido et al. 2003), hybrid zone problems for (Alexander 1962, Walker dynamics (Howard 1986, Howard and Waring 1991, 1964). Frequently among these groups, new species Howard et al. 1993, Chu et al. 1995), competition and are recognized not by museum specimens but rather habitat selection (Howard and Harrison 1984a, by detailed ecological and behavioral observations. 1984b), and reproductive protein evolution (W.E.B., For example, crickets of the genus Allonemobius unpublished data). Their role as important study sys- (Hebard) were initially constrained to a single species tems may be because of their abundance. Indeed, (DeGeer 1773; Alexander and Thomas 1959, Vickery these species have been studied most intensively along and Johnstone 1970). Intense study by Fulton (1931, the eastern coast of North America where they are 1933, 1937) revealed physiological distinctions be- perhaps the most abundant singing orthopteran. How- tween populations that lead to the eventual descrip- ever, most species in this genus are rarely encountered tion of several unique species (Alexander and Thomas and relatively unknown. The majority of work on this 1959). Despite these intense ecological, behavioral, genus has focused on the three species exhibiting the and morphological studies, additional cryptic species highest abundance on the east coast: Allonemobius were found after detailed genetic work (Howard 1983, Howard and Furth 1986). fasciatus (De Geer), (Scudder), All members of the genus Allonemobius are small, and (Alexander & Thomas). Un- ground-dwelling omnivores with most species dis- fortunately, other members of this genus remain in tributed broadly across eastern North America. They their shadow. have become well known through their role in the Although little work has been done outside of east- description of major patterns and processes in ecology ern North America, the presence of geographic bar- and evolution. Species in this genus have been used to riers and diverse habitats in western North America, elucidate the process of speciation (Gregory and suggests that there may be many undiscovered Howard 1993, 1994; Gregory et al. 1998; Doherty and isolated populations in the west. We know of only Howard 1996; Veech et al. 1996; Howard et al. 1998), two taxonomic studies of Allonemobius in the west; sexual selection (Howard and Gregory 1993; Gregory Vickery and Johnstone (1970) collected extensively and Howard 1996; Fedorka and Mousseau 2002a,b), across Canada, and Weibel (1996; see also Weibel and sexual conßict (Fedorka and Mousseau 2004), mating Howard 2000a,b) sampled Allonemobius across the system evolution (Sadowski et al. 2002), genetic dif- western United States. Weibel (1996) discovered sev- eral new populations that she suggested were com- 1 Corresponding author, e-mail: [email protected]. posed of new species. Her work suggests that at least

0013-8746/06/0449Ð0456$04.00/0 ᭧ 2006 Entomological Society of America 450 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 99, no. 3 some species of Allonemobius remain to be discovered. Table 1. Allele frequency of new populations with number of Herein, we describe one such species. individuals in parentheses

Locus Allele Salado, TX Atoka, OK Materials and Methods MDH 1.6 0.9 (9) 1.0 (22) 1.0 0.1 (9) In August1999, W.E.B. collected crickets at the Con- HK 1.0 1.0 (8) 1.0 (15) federate Memorial Museum and rest stop along U.S. IDH 1.5 0.9 (9) 1.0 (21) Hwy. 69 in Atoka, OK, and south of Salado, TX, at mile 1.0 0.1 (9) marker 283 on Interstate Hwy. 35. The Atoka collec- ME 1.0 1.0 (9) 1.0 (22) AAT (GOT) 2.0 0 0.14 (18) tion was transported back to the laboratory alive and 1.0 0.69 (7) 0.86 (18) used for calling song, genetic, and morphological 0.8 0.31 (7) 0 analyses. The Salado collection was immediately fro- PGI 1.6 0 0.02 (22) zen in liquid nitrogen, transported back to the labo- 1.17 0.5 (1) 0.93 (22) Ϫ Њ 1.0 0.5 (1) 0.05 (22) ratory, and transferred to 80 C until used for genetic EST-1 1.1 0.13 (8) analyses. 1.0 0.5 (8) Genetic analysis of individual crickets was performed 0.9 0.31 (8) using allozyme electrophoresis. Electrophoretic analysis 0.8 0.06 (8) G-PGD 1.0 1.0 (8) 1.0 (22) was based on 14 enzyme loci, including isocitrate de- PGM 1.0 1.0 (9) 1.0 (22) hydrogenase (IDH), malate dehydrogenase (MDH), MPI 1.0 1.0 (9) hexokinase (HK), aspartate aminotransferase (AAT), G6PD 1.0 1.0 (1) esterase-1 (EST-1), phosphoglucose isomerase (PGI), ACP 1.0 1.0 (5) ␣-GPD 1.0 1.0 (5) 6-phosphogluconate dehydrogenase (6-PGD), malic en- GP 1.0 1.0 (5) zyme (ME), phosphogucomutase (PGM), mannose-6- phosphate isomerase (MPI), acid phosphatase (ACP), Allele names are based on the starch electrophoretic mobility of the ␣-glycerophosphate dehydrogenase (␣-GPD), glucose- allele relative to that of the most common allele in A. fasciatus. 6-phosphate dehydrogenase (G6PD), and general pro- tein (GP). A single hind leg was removed from each cricket and analyzed as in Howard (1982). these analyses included both the electrophoretic data Morphological analyses were performed on crickets mentioned above as well as data previously published from Atoka that had been previously identiÞed by by Howard (1983). Pairwise genetic distances were electrophoretic and calling song analysis. Eight mor- estimated using the methods of Nei (1972), Cavalli- phological traits were measured, including shared and Sforza and Edwards (1967), and Reynolds et al. sex-speciÞc characters. These characters were chosen (1983). Pairwise distances were used to reconstruct to match those used by Howard and Furth (1986) in phylogenetic relationships by using both least squares describing other members of the genus. For both and minimum evolution methods with and without the sexes, we measured the body length (Lb), pronotum assumption of a molecular clock in the FITCH and length (Lp), pronotum width (Wp), and hind femur KITCH programs. Neighbor-joining and unweighted length (Lf). For males, we measured characters of the pair-group methods with arithmetic average methods tegmen as these characters may affect calling song. were used with the NEIGHBOR program. Data sets Tegmen characters were measured as in Fulton (1931) were resampled 1000 times to generate bootstrap sup- and Howard and Furth (1986) and included the stridu- port using SEQBOOT, and consensus trees were ob- latory vein measured from the mesal origin to the tained using CONSENSE. ulnar vein (Tg1), the distance from the ulnar vein to the lateral border of dorsal tegmen surface (Tg2), and Results the distance from the stridulatory vein to the apex of the tegmen (Tg3). For females, we measured the We genetically analyzed a total of 31 crickets from length of the ovipositor (Lo). the two collections (22 from Atoka, OK, and nine from To record male calling songs, males were individ- Salado, TX) for 14 electrophoretic loci. Most loci are ually isolated in 17- by 12- by 6-cm plastic boxes with invariant and have alleles typical of the A. fasciatus mesh covers. Songs were recorded using a Realistic species group (Table 1). However, two allozyme loci Cardioid dynamic microphone and a Sony MZ-R50 are Þxed for novel alleles in one population. These recorder at 22Ð24ЊC. Songs were analyzed using the novel alleles occur at high frequencies in the other Canary software (Cornell Laboratory of Ornithology, population. A third locus exhibits variation in both Ithaca, NY; Charif et al. 1995). By isolating individual populations, with high frequency alleles that are rare males, we ensured that songs produced were not in within the A. fasciatus group. In the Atoka population, response to stimulus from females or other males; as a the allele for the MDH enzyme is Þxed and migrates result, the songs recorded are calling songs and not at 1.6 times the distance as the most common allele courtship or aggressive songs. in A. fasciatus. Similarly, the IDH locus is Þxed for a Phylogenetic analyses were performed to assess the novel allele in the Atoka population, which migrates relationship of newly collected crickets to the remain- 1.5 times the distance as the most common allele in ing members the genus Allonemobius by using pro- A. fasciatus. The PGI locus was only well sampled in grams in PHYLIP (Felsenstein 2004). The data used in the Atoka population. Here, an allele migrating 1.17 May 2006 BRASWELL ET AL.: NEW CRYPTIC SPECIES OF Allonemobius 451

times the distance as the most common allele in the A. fasciatus group was common. The Salado population, however, seems to consist of two genotypic classes along with potential advanced backcross individuals. The Salado collection has one individual that seems to be a member of a recently discovered yet undescribed species, A. sp. nov. TX (Traylor et al., unpublished data; Marshall 2004). This Fulton (1930) Hebard (1913) Hebard (1913) individual is homozygous at IDH, MDH, and AAT for the 1.3, 1.0, and 2.0 alleles, respectively. The other eight individuals from Salado are homozygous for the same novel alleles found in Atoka at the MDH and HK loci (1.6 and 1.0, respectively). Six of these eight in- dividuals are also homozygous for the novel IDH allele found in Atoka (namely, the 1.5 allele). However, the remaining two individuals are heterozygous at IDH (carrying both the 1.5 and 1.3 alleles). That is, all

A. shalontaki. individuals in Salado are genotypically identical to the crickets from Atoka except three crickets. One of these three is a member of A. sp. nov. TX. The re- maining two crickets are genotypically identical to Atoka crickets at all loci except IDH where they are heterozygous for the Atoka allele and the A. sp. nov. TX allele. We measured morphological characters of a sample of individuals from Atoka that were previously iden- tiÞed by song analysis and genotyped by electro- phoretic analysis. Morphological data are presented in Table 2 (and in Diagnosis below). The clear pattern is that these crickets are larger than A. fasciatus or A. socius (Table 2). Other than body size, these crick- ets seem similar to A. fasciatus and A. socius, including similar head coloration, one of the most variable char- acters in this genus of cryptic species. Calling songs were recorded from 18 Þeld-collected crickets from the Atoka population. These crickets exhibited a unique male calling song that lies outside the range of variation of all known Allonemobius spe- cies. The song is characterized as a relatively long chirp (or short trill) that crescendos through the Allonemobius length of the chirp (Fig. 1). That is, the chirp lasts between 720 and 770 ms with a dominant frequency of 8 kHz and exhibits a unique pattern of increasing amplitude (sound pressure) through the chirp (Fig. 1B and C). Chirps are regularly timed occurring Ϸ0.6 s apart (Fig. 1A) and are composed of Ϸ50 pulses. Although chirps are regularly timed, the song is com- plex at the level of the chirp. Individual chirps have a temporal structure as there is variation in the rate of the crescendo within chirps (Fig. 1B). SpeciÞcally, the amplitude increases 22% by 0.4 s into the song, and another 60% by the end of the song. The song is further 11) 10.8 (10Ð12) 2.3 (2.1Ð2.5) 3.4 (3.0Ð3.8) 7.5 (7.0Ð8.0) 7.6 (7.0Ð8.5) complicated by variation in pulse duration and pulse 7) 11.9 (11Ð13) 2.1 (1.9Ð2.2) 3.2 (2.7Ð3.7) 6.9 (6.0Ð7.5) 1.0 (8.0Ð11) 1.4 (1.3Ð1.7) 3.9 (3.0Ð4.6) This study ϭ n

ϭ rate. The Þnal 10 to 11 pulses of each chirp are longer

n and occur at a slower rate than the Þrst 40 pulses. Moreover, these last pulses vary little in amplitude Male ( Female ( FemaleMaleFemaleMaleFemale - 7.4 8.3 7.2 8.1 2.4 1.8 2.0 1.6 1.8 3.4 2.5 2.6 2.1 2.3 7.7 5.7 6.3 5.2 6.3 6.8 6.2 8.0 MaleFemaleMaleFemaleMaleFemaleMaleFemaleMale 10.0 9.1FemaleMale 10.7 9.8Female 11.0Male 10.9 2.1 1.8 8.7 9.4 2.1 1.8 10.6 11.2 1.8 2.1 10.9 9.7 2.9 2.7 1.8 2.1such - 2.9 2.6 2.2 2.5 2.7 2.9 that 2.3 1.9 6.8 6.4 2.7 2.9 the 6.9 2.2 6.3 3.2 3.4 additional 6.4 7.1 3.1 2.8 6.9 6.3 7.1 7.3 7.5 - 8.3 60% 8.1 7.9 6.8 increase - 8.2 1.0 9.4 7.1 0.9in 8.6 1.1 amplitude 1.1 1.2 1.1 1.3 1.2 1.0 3.5 1.2 3.5 1.4 3.8 Howard and Furth 1.2 (1986) Howard and 3.5 Furth (1986) Howard and 4.5 Furth (1986) Howard and 3.7 Furth (1986) Howard and Furth (1986) Howard and Furth (1986) occurs before these last pulses (Fig. 1B). The initial 40 pulses have an average pulse rate of 69 pulses per second, whereas the Þnal pulses have an average pulse Table 2. Summary of morphological characteristics in the genus

Species Sex Lb Lp Wprate of Lf 45.6 pulses Lo per second. Tg1 Finally, Tg2 the last ten Tg3 Citation Data are presented as mean values along with citations from which the data are drawn. The range of variation is presented in parentheses for A. shalontaki A. maculatus A. griseus A. socius A. fasciatus A. allardi A. tinnulus A. walkeri A. fultoni A. sparsalsus pulses exhibit frequency modulation with a rapid rise 452 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 99, no. 3

Fig. 1. Representative male calling song of A. shalontaki from Atoka, OK, recorded at 22ЊC. (A). Waveform depicting 9 s of male calling song. (B). Detailed waveform view of one chirp (or trill) lasting Ͻ0.8 s and depicting a crescendo in amplitude through the course of the chirp as well as a decrease in pulse rate toward the end of the chirp. (C). Spectrogram of the same chirp as in B, indicating the dominant frequency of 8 kHz. followed by a slower decline in frequency not detect- lected by W.E.B. from same area as holotype are able within the initial 40 pulses (Fig. 1C). housed in the Florida State Collection of and in the New Mexico State University entomology collection (accession nos. 26267Ð26274). Allonemobius shalontaki Braswell, n. sp. Color and Pattern. Head with dark longitudinal Choctaw ground cricket stripes from base onto vertex between eyes, as in Type Data. HOLOTYPE: 1 & was collected by A. fasciatus, and A. socius. W.E.B, VIII-1999 at Confederate Memorial Museum Morphometrics. Morphological measurements (in and rest stop along U.S. Hwy. 69 in Atoka, OK, USA, millimeters) are presented in Table 2. Average body and is deposited in the Florida State Collection of length is greater in males (11.9 mm) than in females Arthropods. PARATYPES: Additional specimens col- (10.8 mm). However, females are larger in all other May 2006 BRASWELL ET AL.: NEW CRYPTIC SPECIES OF Allonemobius 453 measurements, including pronotum length and width, A. fasciatus clade, and the MDH allele is novel to the and femur length (Table 2). entire genus (Table 1). Although the lack of geno- Biological Notes. A. shalontaki was Þrst collected typically distinct sympatric congeners poses a theo- within the Choctaw Nation at Atoka, OK. Therefore, retical problem to the utility of genetic divergence as the species is named in honor of the Choctaw Nation. a sign of reproductive isolation, extensive work has The speciÞc epithet shalontaki means “cricket” in the shown this pattern to be a robust method for identi- Choctaw language. fying cryptic species (Howard and Furth 1986, Like many of its congeners in eastern North Amer- Howard et al. 1998). ica, A. shalontaki is an inhabitant of short grasslands. Although allozyme divergence represents persua- It is likely a more drought-tolerant species than its sive data supporting species status of A. shalontaki, the congeners; however, the two localities in which this extreme divergence in calling song makes explana- cricket has been found were watered lawns. tions other than reproductive isolation unlikely. Al- though studies of female calling song preference An Electrophoretic Key to Allonemobius Species, have not been performed, it is unlikely that such di- by Using A. fasciatus as a Standard vergent calling songs (let alone electrophoretic vari- ants) could remain intact if these crickets were ex- 1. MDH migrates the same distance as standard . 3 changing genes with any known Allonemobius species. MDH migrates farther than standard ...... 2 In addition to supporting species status, male calling 2. MDH migrates 1.3 times the distance of stan- song likely acts as a prezygotic isolating barrier. Al- dard...... 4 though mating studies have not been performed to MDH migrates 1.6 times the distance of stan- conÞrm reproductive isolation, morphological, elec- dard ...... A. shalontaki trophoretic, and calling song analyses all indicate a 3. HK migrates the same distance as standard .... lack of gene ßow with previously described species. As ...... A. fasciatus a result, we feel conÞdent in naming A. shalontaki as HK migrates 0.8 times the distance of the standard. a new species...... A. socius Electrophoretic data suggest two possibilities re- 4. ME migrates the same distance as standard . . 5 garding the Salado population. Either A. shalontaki ME migrates faster or slower than standard . . 8 exhibits allelic variation at the IDH locus and is not 5. IDH migrates at least 1.8 times the distance of the Þxed for the 1.5 allele as suggested by the Atoka pop- standard ...... 6 ulation, or A. shalontaki and A. sp. nov. TX have ex- IDH migrates no Ͼ1.5 times the distance of stan- changed genes at sometime in the, perhaps recent, dard ...... 7 past. Although these two possibilities are not mutually 6. HK migrates the same distance as standard . . . exclusive, indirect evidence does make some predic- ...... A. allardi HK migrates 1.3 times the distance of standard . . tions. Given that these species co-occur in Salado, it ...... A. tinnulus seems reasonable to suggest that at least limited hy- 7. HK migrates the same distance as standard . . . bridization is or has occurred between these two spe- ...... A. fultoni cies. Additional work is required to differentiate be- HK migrates 1.3 times the distance of standard . . tween these two possibilities...... A. walkeri In an attempt to understand the evolutionary rela- 8. ME migrates slower than standard...... tionships within the genus Allonemobius, we recon- ...... A. maculatus structed phylogeny by using genetic distance esti- ME migrates faster than standard . . A. griseus mated from electrophoretic data (Table 3). We found that tree topology was resilient to both the method of reconstruction and genetic distance estimators. Most Discussion methods revealed A. shalontaki as basal to the A. fas- We present data supporting species status of the ciatus–A. socius clade (Fig. 2). Indeed, only two of crickets we have named A. shalontaki. The data sets we the 18 methods, least squares (with a molecular present are of varying importance in supporting spe- clock) and unweighted pair-group method with arith- cies status. For example, morphological data suggest metic average methods by using Reynolds genetic that although A. shalontaki is larger than other mem- distance, placed A. shalontaki differently. These two bers of the A. fasciatus species group, it is generally methods place A. shalontaki basal to both the A. fas- indistinguishable from these species. However, be- ciatus and A. allardi clades, e.g., (((((fasciatus, socius), cause body size measurements of other Allonemobius (((allardi, tinnulus), walkeri), fultoni)), shalontaki), species are from published results rather than sympa- griseus), maculatus). Although all of these analyses are tric, or even western, populations it is impossible to based on a small data set and exhibit moderate boot- know whether size variation is because of divergence strap support, they demonstrate the interspeciÞc re- or environmental variation. lationships within the genus and are largely consistent Electrophoretic data, however, suggest that A. sha- with previous work on the genus (Howard 1983). lontaki represents a unique gene pool from other Thus, these analyses pose viable hypotheses regarding Allonemobius species. At a minimum, the IDH locus in the placement of A. shalontaki. Additional data are A. shalontaki is Þxed for an allele unknown within the required to assess the evolutionary relationships 454 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 99, no. 3

Table 3. Summary of pairwise genetic distance among Allonemobius species by using nine loci

shalontaki allardi tinnulus fultoni walkeri fasciatus socius maculatus griseus shalontaki *** 0.602 0.600 0.400 0.508 0.406 0.580 1.200 0.765 allardi 0.376 *** 0.129 0.203 0.402 0.608 0.447 0.953 0.767 tinnulus 0.371 0.104 *** 0.343 0.400 0.606 0.466 0.824 0.765 fultoni 0.251 0.089 0.206 *** 0.308 0.606 0.608 1.000 0.565 walkeri 0.352 0.203 0.198 0.203 *** 0.606 0.608 1.000 0.673 fasciatus 0.254 0.380 0.375 0.409 0.390 *** 0.325 1.000 0.971 socius 0.367 0.292 0.285 0.370 0.351 0.204 *** 1.123 0.974 maculatus 1.083 0.756 0.585 0.796 0.777 0.789 1.008 *** 1 griseus 0.579 0.550 0.544 0.397 0.519 0.809 0.771 0.793 ***

NeiÕs distance is presented below the diagonal and CavalliÐSforzaÕs distance is presented above the diagonal. within this genus and such analyses, based on DNA fore, it seems that the lack of recognition of this spe- sequence, are underway (W.E.B., unpublished). cies is because of a lack of work in the west rather than Although most species in the genus Allonemobius its cryptic nature. IdentiÞcation of A. shalontaki and an are morphologically cryptic, many were recognized to as yet undescribed species from Texas (Traylor et al., be “physiologically distinct” (Fulton 1931, 1933, 1937) unpublished data; Marshall 2004) along with two po- before their eventual description as distinct species tential new species in the Dakotas (Weibel 1996) (Alexander and Thomas 1959). Unlike these species, suggest that the genus Allonemobius may be much A. shalontaki was not previously recognized. Although more diverse than currently recognized. It seems that A. shalontaki cannot be identiÞed morphologically, we have only begun to sample the diversity of this males can be easily identiÞed by calling song. There- genus and additional work in the west may be fruitful.

Fig. 2. Unweighted pair-group method with arithmetic average tree based on CavalliÐSforzaÕs estimate of genetic distance. Values at nodes are percentage of bootstrap support based on 1,000 replicates. May 2006 BRASWELL ET AL.: NEW CRYPTIC SPECIES OF Allonemobius 455

Acknowledgments Fulton, B. B. 1937. Experimental crossing of subspecies in fasciatus (Orthoptera). Ann. Entomol. Soc. We thank William Braswell for help collecting and sug- Am. 30: 201Ð207. gesting the initial collection site, without whose help we Gregory, P. G., and D. J. Howard. 1993. Laboratory hybrid- would not have found this species. T. J. Walker provided ization studies of and A. socius helpful discussions regarding species status and song analysis (Orthoptera: Gryllidae). Ann. Entomol. Soc. Am. 86: 694Ð and signiÞcantly improved the manuscript. Funding was pro- 701. vided by Sigma Xi, New Mexico State University Biology Departmental Fellowship, and the Environmental Protection Gregory, P. G., and D. J. Howard. 1994. A post-insemination Agency STAR/GRO Fellowship awarded to W.E.B. and barrier to fertilization isolates two closely related ground L.M.B. as well as the OrthopteristsÕ Society and University of crickets. Evolution 48: 705Ð710. Virginia Blandy Experimental Farm Graduate Fellowship Gregory, P. G., and D. J. Howard. 1996. Multiple mating in awarded to W.E.B. D.J.H. was supported by National Science natural populations of ground crickets. Entomol. Exp. Foundation DEB 0316194 and IRCEB 0111613. Appl. 78: 355Ð356. Gregory, P. G., M. D. Remmenga, and D. J. Howard. 1998. Patterns of mating between two closely related ground crickets are not inßuenced by sympatry. Entomol. Exp. References Cited Appl. 87: 263Ð270. Hebard, M. 1913. A revision of the genus Nemobius Alexander, R. D. 1962. The role of behavioral study in (Orthoptera: Gryllidae) found in North America north of cricket classiÞcation. Syst. Zool. 11: 53Ð72. the Isthmus of Panama. Proc. Acad. Natl. Sci. Phila. 65: Alexander, R. D., and E. S. Thomas. 1959. Systematic and behavioral studies on the crickets of the Nemobius fas- 394Ð492. ciatus group (Orthoptera: Gryllidae: Nemobiinae). Ann. Howard, D. J. 1982. Speciation and coexistence in a group Entomol. Soc. Am. 52: 591Ð605. of closely related ground crickets. Ph.D. dissertation, Yale Bradford, M. J., and D. A. Roff. 1993. Bet hedging and the University, New Haven, CT. diapause strategies of the cricket Allonemobius fasciatus. Howard, D. J. 1983. Electrophoretic survey of eastern Ecology 74: 1129Ð1135. North American Allonemobius (Orthoptera: Gryllidae): Carrie`re, Y., S. Masaki, and D. A. Roff. 1997. The coadap- evolutionary relationships and the discovery of three new tation of female morphology and offspring size: a com- species. Ann. Entomol. Soc. Am. 76: 1014Ð1021. parative analysis in crickets. Oecologia 110: 197Ð204. Howard, D. J. 1986. A zone of overlap and hybridization Cavalli-Sforza, L. L., and A.W.F. Edwards. 1967. Phyloge- between two ground cricket species. Evolution 40: 34Ð43. netic analysis: analysis and methods. Evolution 32: 550Ð Howard, D. J., and D. G. Furth. 1986. Review of the 570. Allonemobius fasciatus (Orthoptera: Gryllidae) com- Charif, R. A., S. Mitchell, and C. W. Clark. 1995. Canary 1.12 plex with the description of two new species separated userÕs manual. Cornell Laboratory of Ornithology, Ithaca, by electrophoresis, songs, and morphometrics. Ann. NY. Entomol. Soc. Am. 79: 472Ð481. Chu, J., and D. J. Howard. 1998. Genetic linkage maps of the Howard, D. J., and P. G. Gregory. 1993. Post-insemination ground crickets Allonemobius fasciatus and A. socius using signaling systems and reinforcement. Phil. Trans. R. Soc. RAPD and allozyme markers. Genome 41: 841Ð847. Lond. B 340: 231Ð236. Chu, J., E. Powers, and D. J. Howard. 1995. Gene exchange Howard, D. J., and R. G. Harrison. 1984a. Habitat segrega- in a cricket hybrid zone. J. Hered. 86: 17Ð21. tion in ground crickets: experimental studies of adult DeGeer, C. 1773. Memoires pour servir a lÕhistoire des in- survival, reproductive success, and oviposition prefer- sects, III. Hosselberg, Stockholm, Sweden. ence. Ecology 65: 61Ð68. Doherty, J. A., and D. J. Howard. 1996. Lack of preference Howard, D. J., and R. G. Harrison 1984b. Habitat segrega- for conspeciÞc calling songs in female crickets. Anim. tion in ground crickets: the role of interspeciÞc compe- Behav. 51: 981Ð990. tition and habitat selection. Ecology 65: 69Ð76. Fedorka, K. M., and T. A. Mousseau. 2002a. Material and Howard, D. J., and G. L. Waring. 1991. Topographic diver- genetic beneÞts of female multiple mating and polyandry. sity, zone width, and the strength of reproductive isola- Anim. Behav. 64: 361Ð367. tion in a zone of overlap and hybridization. Evolution 45: Fedorka, K. M., and T. A. Mousseau. 2002b. Tibial spur feed- 1120Ð1135. ing in ground crickets: larger males contribute larger gifts Howard, D. J., P. G. Gregory, J. Chu, and M. L. Cain. 1998. (Orthoptera: Gryllidae). Fla. Entomol. 85: 317Ð323. ConspeciÞc sperm precedence is an effective barrier to Fedorka, K. M., and T. A. Mousseau. 2004. Female mating hybridization between closely related species. Evolution bias results in conßicting sex-speciÞc offspring Þtness. Nature (Lond.) 429: 65Ð67. 52: 511Ð516. Felsenstein, J. 2004. PHYLIP (phylogenetic inference pack- Howard, D. J., M. Reece, P. G. Gregory, J. Chu, and age) version 3.6. Distributed by author. Department of M. L. Cain. 1998. The evolution of barriers to fertiliza- Genome Sciences, University of Washington, Seattle, tion between closely related organisms, pp. 279Ð288. In WA. D. J. Howard and S. H. Berlocher [eds.], Endless forms: Fulton, B. B. 1930. A new species of Nemobius from North species and speciation. Oxford University Press, New Carolina (Orthoptera: Gryllidae). Entomol. News 41: 38Ð York. 42. Howard, D. J., G. L. Waring, C. A. Tibbets, and P. G. Gregory. Fulton, B. B. 1931. A study of the genus Nemobius 1993. Survival of hybrids in a mosaic hybrid zone. Evo- (Orthoptera: Gryllidae). Ann. Entomol. Soc. Am. 24: 205Ð lution 47: 789Ð800. 237. Marshall, J. L. 2004. The Allonemobius-Wolbachia host-en- Fulton, B. B. 1933. Inheritance of song in hybrids of two dosymbiont system: evidence for rapid speciation and subspecies of Nemobius fasciatus (Orthoptera). Ann. against reproductive isolation driven by cytoplasmic in- Entomol. Soc. Am. 26: 368Ð376. compatibility. Evolution 58: 2409Ð2425. 456 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 99, no. 3

Mousseau, T. A., and D. J. Howard. 1998. Genetic variation Tanaka, S., and V. J. Brookes. 1983. Altitudinal adaptation of in cricket calling song across a hybrid zone between two the life cycle in Allonemobius fasciatus. Can. J. Zool. 61: sibling species. Evolution 52: 1104Ð1110. 1986Ð1990. Nei, M. 1972. Genetic distance between populations. Am. Veech, J. A., J. H. Benedix, Jr., and D. J. Howard. 1996. Lack Nat. 106: 283Ð292. of calling song displacement between two closely related Olvido, A. E., E. S. Elvington, and T. A. Mousseau. 2003. ground crickets. Evolution 50: 1982Ð1989. Relative effects of climate and crowding on wing poly- Vickery, V. R., and D. E. Johnstone. 1970. Generic status of morphism in the southern ground cricket, Allonemobius some Nemobiinae (Orthoptera: Gryllidae) in northern socius (Orthoptera: Gryllidae). Fla. Entomol. 86: 158Ð164. North America. Ann. Entomol. Soc. Am. 63: 1740Ð1749. Reynolds, J. B., B. S. Weir, and C. C. Cockerham. 1983. Walker, T. J. 1964. Cryptic species among sound-producing Estimation of coancestry coefÞcient: basis for a short- term genetic distance. Genetics 105: 767Ð779. ensiferan Orthoptera (Gryllidae and Tettigoniidae). Roff, D. A., and M. J. Bradford. 2000. A quantitative genetic Q. Rev. Biol. 39: 345Ð355. analysis of phenotypic plasticity of diapause induction in Weibel, A. C. 1996. Distribution, population genetics and the cricket Allonemobius socius. Heredity 84: 193Ð200. evolutionary relationships of the striped ground cricket Roff, D. A., and T. A. Mousseau. 1999. Does natural selection (genus Allonemobius) across North America. M.S. thesis, alter genetic architecture? An evaluation of quantitative New Mexico State University, Las Cruces, NM. genetic variation among populations of Allonemobius so- Weibel, A. C., and D. J. Howard. 2000a. Discovery of new cius and A. fasciatus. J. Evol. Biol. 12: 361Ð369. populations of striped ground crickets in western North Roff, D. A., T. A. Mousseau, and D. J. Howard. 1999. Vari- America: distribution, subdivision, and introduction de- ation in genetic architecture of calling song among pop- tected by allozyme data. J. Entomol. Res. Soc. 2: 1Ð25. ulations of Allonemobius socius, A. fasciatus, and a hybrid Weibel, A. C., and D. J. Howard. 2000b. Canonical corre- population: drift or selection? Evolution 53: 216Ð224. lation of allozyme and morphological variation across Sadowski, J. A., J. L. Grace, and A. J. Moore. 2002. Complex distinct geological features in one species of striped courtship behavior in the striped ground cricket Al- ground crickets. J. Entomol. Res. Soc. 2: 15Ð29. lonemobius socius (Orthoptera: Gryllidae): does social environment affect male and female behavior? J. Behav. 15: 69Ð84. Received 29 September 2005; accepted 6 December 2005.